POLITECNICO DI TORINO
DIPARTIMENTO DI INGEGNERIA MECCANICA E AEROSPAZIALE
Master Degree Thesis
Euro NCAP 2020: the evolution of partial overlap frontal tests and its
impact on effective vehicle safety
A simulation analysis of the Euro NCAP ODB and MPDB procedures
Principal Supervisor
prof. Andrea Tonoli
Secondary Supervisors:
prof. Gianpiero Mastinu prof. Marco Anghileri
Candidate
Tommaso Maria Verri
Industrial Supervisor Pininfarina Engineering
ing. Francesco Macheda
In the last two decades, consumer testing programmes such as Euro NCAP have driven a substantial improvement in vehicle safety thanks to their effect on public opinion, which was in turn seen as a marketing opportunity by manufacturers. Nonetheless, in the EU, road accidents are still a major cause of death, with over 25 000 people killed every year and a much greater number suffering serious injuries. Out of this number, a substantial part is caused by two car frontal impacts, which are currently tested both by legislation and Euro NCAP with the Offset Deformable Barrier test. However, it has been historically pointed out that this test is not closely representative of real life scenarios as it cannot assess the negative impacts of poor compatibility between vehicle masses, structural designs and front end rigidities. For this reason, a new test procedure has been devised and will be introduced in 2020. The Mobile offset Deformable Barrier test is based on the use of a trolley of a set mass representing the average of the european circulating fleet, which will impact against a vehicle moving at equal velocity in opposite direction. The deformable element used will also allow the assessment of partner protection criteria based on its own deformation. This change represents a new challenge for OEMs and design firms, as it will increase the level of complexity required to design frontal structures, in order for them to guarantee the same level of performance shown in the ODB.
Hence, the aim of this study is to gain an initial understanding of the effects that the new test has on vehicle passive safety by performing comparative simulations between the two procedures on a number of full vehicle mathematical models. The four vehicles to be included were selected in order to have a range of different masses and vehicle design philosophies. Initially, a correlation study based on full width rigid barrier tests was per- formed in order to understand the level of representativeness of the models. This showed that two out of the four were closely related to physical performance, while other two presented a level of discrepancy. Next, the main part of the study involved the simulation of the vehicles in the ODB and MPDB procedures, which resulted in an in depth anal- ysis completed in terms of crash pulse, section forces, structural deformation and cabin intrusion. The comparative study highlighted an increased harshness of the MPDB test due to its reduced timeframe, in all four vehicles. At the same time it showed a clear correlation between performance and mass: the lighter vehicle underwent substantially higher damage, the model with mass similar to the barrier only minor negative effects, while the two vehicles with heavier mass saw a clear improvement in their deformation, intrusion and acceleration. Furthermore, the deformation of the barrier and the dynamic data of the trolley allowed to assess the partner protection level of the models. The results showed that the performance was closely related to chassis design: the very light and very heavy vehicles achieved comparably negative results, while the vehicle with more advanced frontal structures proved to be substantially better. In conclusion, the study highlighted the very poor results that can currently be obtained when using a ladder chassis design.
List of Tables vii
List of Figures viii
1 Introduction 1
1.1 The issue of representative crash testing . . . 1
1.2 Aims and objectives . . . 2
1.3 Report structure . . . 3
2 State of the art 5 2.1 The evolution of Euro NCAP . . . 5
2.1.1 Euro NCAP future Roadmaps . . . 9
2.2 From ODB to MPDB . . . 10
2.2.1 The Offset Deformable Barrier test . . . 10
2.2.2 The Mobile Offset Deformable Barrier test . . . 15
2.3 Crash test simulation in the vehicle design process . . . 20
3 Methodology 23 3.1 Simulation softwares . . . 23
3.2 Vehicle models . . . 25
3.2.1 Toyota Yaris. . . 26
3.2.2 Honda Accord . . . 30
3.2.3 Chevrolet Silverado . . . 33
3.2.4 U Model . . . 38
3.3 Barrier models . . . 41
3.4 Schedule of conducted tests . . . 42
4 Correlation Study 45 4.1 The issue of correlating physical crash testing with mathematical models . 45 4.2 Correlation methodology . . . 46
4.3 Toyota Yaris . . . 47
4.4 Honda Accord . . . 53
4.5 Chevrolet Silverado . . . 60
4.6 U Model . . . 65
4.7 Conclusions . . . 68
5 Results 71
5.1 Toyota Yaris . . . 71
5.1.1 ODB test . . . 71
5.1.2 MPDB test . . . 78
5.1.3 Comparison . . . 84
5.2 Honda Accord . . . 93
5.2.1 ODB test . . . 93
5.2.2 MPDB test . . . 98
5.2.3 Comparison . . . 102
5.3 Chevrolet Silverado . . . 106
5.3.1 ODB test . . . 106
5.3.2 MPDB test . . . 113
5.3.3 Comparison . . . 117
5.4 U Model . . . 124
5.4.1 ODB test . . . 124
5.4.2 MPDB test . . . 130
5.4.3 Comparison . . . 136
5.5 Conclusions . . . 145
6 Partner protection analysis 147 6.1 Partner protection in Euro NCAP 2020 . . . 147
6.2 Analysis methodology in simulation . . . 148
6.3 Partner protection results . . . 149
6.3.1 Toyota Yaris. . . 149
6.3.2 Honda Accord . . . 152
6.3.3 Chevrolet Silverado . . . 155
6.3.4 U Model . . . 158
6.4 Conclusions . . . 161
7 Conclusions and further work 163 7.1 Simulation study conclusions. . . 163
7.2 Further work . . . 166
Bibliography i
A Simulation energy charts v
B Correlation study photographs ix
2.1 ODB test specification . . . 13
2.2 ODB construction specifications . . . 14
2.3 MPDB test specification . . . 18
2.4 PDB construction specifications . . . 19
3.1 2009 Toyota Yaris specifications . . . 26
3.2 2011 Honda Accord specifications . . . 30
3.3 2014 Chevrolet Silverado specifications . . . 34
3.4 Tests schedule . . . 43
4.1 US-NCAP - utilised intrusion measurement points [1] . . . 46
4.2 Toyota Yaris - full width rigid barrier dynamic data . . . 48
4.3 Toyota Yaris - full width rigid barrier deformation measurements [mm] . . 51
4.4 Honda Accord - full width rigid barrier dynamic data . . . 55
4.5 Honda Accord - full width rigid barrier deformation measurements [mm] . 58 4.6 Chevrolet Silverado - full width rigid barrier dynamic data . . . 61
4.7 Chevrolet Silverado - full width rigid barrier deformation measurements [mm] 63 4.8 U Model - ODB 56km/h intrusion measurements [2]. . . 67
5.1 Toyota Yaris ODB - intrusion measurements . . . 75
5.2 Toyota Yaris MPDB - intrusion measurements . . . 82
5.3 Toyota Yaris - ODB vs MPDB dynamic data . . . 86
5.4 Toyota Yaris - intrusion measurement comparison . . . 90
5.5 Honda Accord ODB - intrusion measurements . . . 97
5.6 Honda Accord MPDB - intrusion measurements . . . 101
5.7 Honda Accord - intrusion measurement comparison . . . 106
5.8 Chevrolet Silverado ODB - intrusion measurements . . . 112
5.9 Chevrolet Silverado MPDB - intrusion measurements . . . 117
5.10 Chevrolet Silverado - ODB vs MPDB dynamic data . . . 118
5.11 Chevrolet Silverado - intrusion measurement comparison . . . 123
5.12 U Model ODB - intrusion measurements . . . 130
5.13 U Model MPDB - intrusion measurements . . . 136
5.14 U model - ODB vs MPDB dynamic data . . . 138
5.15 U Model - intrusion measurement comparison . . . 142
List of Figures
2.1 Progress of adult occupant star rating 1997-2007 [3] . . . 7
2.2 Progress of pedestrian protection star rating 1997-2007 [3] . . . 8
2.3 Euro NCAP future Roadmaps . . . 9
2.4 ODB Normal element with bumper [4] . . . 11
2.5 Car-to-car frontal impact speed and serious or fatal casualties [5]. . . 12
2.6 ODB test infographic [6] . . . 13
2.7 Physical model of ODB [7] . . . 13
2.8 MPDB test infographic [9] . . . 17
2.9 PDB physical model [7] and dimensions [10] . . . 19
2.10 Displacement-strength characteristic corridor of Block B [10] . . . 19
2.11 Progressive Deformable Barrier physical model [7] . . . 20
3.1 The simulation process - softwares utilised . . . 24
3.2 2009 Toyota Yaris Sedan - Physical [11] and CAE models . . . 27
3.3 Toyota Yaris engine compartment - Physical [12] and CAE models . . . 28
3.4 Toyota Yaris - passenger compartment details . . . 28
3.5 Toyota Yaris - model details . . . 29
3.6 2011 Honda Accord - Physical [13] and CAE models. . . 31
3.7 Honda Accord engine compartment - Physical [14] and CAE models . . . . 31
3.8 Honda Accord - passenger compartment details . . . 32
3.9 Honda Accord - BIW front section physical [15] and CAE models . . . 32
3.10 Honda Accord - model details . . . 33
3.11 2014 Chevrolet Silverado - Physical [16] and CAE models . . . 34
3.12 Chevrolet Silverado engine compartment - Physical [17] and CAE models . 35 3.13 Chevrolet Silverado - passenger compartment details . . . 35
3.14 Chevrolet Silverado - main ladder frame . . . 36
3.15 Chevrolet Silverado - structure details . . . 37
3.16 U Model - CAE model . . . 38
3.17 U Model details - interior and engine compartment . . . 38
3.18 U Model - ladder type chassis . . . 39
3.19 U Model - details . . . 40
3.20 Offset Deformable Barrier - CAE model . . . 41
3.21 Progressive Deformable Barrier - CAE model. . . 41
3.22 Mobile Offset Deformable Barrier - CAE model . . . 42
4.1 Yaris FWRB - simulation snapshot . . . 48
4.2 Yaris FWRB - acceleration . . . 49
4.5 Yaris FWRB - LHS and RHS comparison. . . 51
4.6 Yaris FWRB - three quarters and underbody comparison . . . 52
4.7 Yaris FWRB - top view comparison . . . 52
4.8 Honda Accord - front structures modification. . . 54
4.9 Accord - comparison of crash pulse before and after modifications . . . 54
4.10 Accord FWRB - simulation snapshot . . . 56
4.11 Accord FWRB - acceleration. . . 56
4.12 Accord FWRB - velocity . . . 57
4.13 Accord FWRB - crush space . . . 57
4.14 Accord FWRB - LHS and RHS comparison. . . 59
4.15 Accord FWRB - top view comparison . . . 59
4.16 Accord FWRB - three quarters and underbody comparison . . . 60
4.17 Silverado FWRB - simulation snapshot . . . 61
4.18 Silverado FWRB - acceleration . . . 62
4.19 Silverado FWRB - velocity . . . 62
4.20 Silverado FWRB - crush space. . . 63
4.21 Silverado FWRB - LHS and RHS comparison . . . 64
4.22 Silverado FWRB - top view comparison . . . 64
4.23 Silverado FWRB - three quarters and underbody comparison . . . 65
4.24 U Model ODB 56km/h - simulation snapshot . . . 67
4.25 U Model ODB 56km/h - RHS and underbody comparison . . . 68
5.1 Toyota Yaris ODB - simulation snapshot . . . 72
5.2 Toyota Yaris ODB - structural collapse detail . . . 73
5.3 Toyota Yaris ODB - acceleration. . . 73
5.4 Toyota Yaris ODB - section forces . . . 74
5.5 Toyota Yaris ODB - firewall intrusion . . . 76
5.6 Toyota Yaris ODB - firewall deformed region . . . 76
5.7 Toyota Yaris ODB - driver door opening deformation . . . 76
5.8 Toyota Yaris ODB - A-pillar deformation . . . 77
5.9 Toyota Yaris ODB - floor and tunnel deformation . . . 77
5.10 Toyota Yaris MPDB - simulation snapshot . . . 79
5.11 Toyota Yaris MPDB - structural collapse detail . . . 79
5.12 Toyota Yaris MPDB - acceleration . . . 80
5.13 Toyota Yaris MPDB - section forces . . . 80
5.14 Toyota Yaris MPDB - firewall intrusion . . . 82
5.15 Toyota Yaris MPDB - firewall deformed region . . . 83
5.16 Toyota Yaris MPDB - floor and tunnel deformation . . . 83
5.17 Toyota Yaris MPDB - driver door opening deformation . . . 84
5.18 Toyota Yaris MPDB - deformed pillar and interior . . . 84
5.19 Toyota Yaris - comparison between ODB and MPDB energy content. . . . 85
5.20 Toyota Yaris - front structures behaviour comparison . . . 86
5.21 Toyota Yaris - crash pulse comparison. . . 87
5.22 Toyota Yaris - velocity trend comparison . . . 87
5.23 Toyota Yaris - section forces comparison . . . 89
5.24 Toyota Yaris - ODB firewall deformation . . . 89
5.25 MPDB . . . 90
5.26 Toyota Yaris - MPDB firewall deformation . . . 90
5.27 Toyota Yaris - interior intrusion comparison . . . 91
5.28 Toyota Yaris - front deformation comparison . . . 92
5.29 Honda Accord ODB- acceleration and section forces . . . 93
5.30 Honda Accord ODB - simulation snapshot . . . 94
5.31 Honda Accord ODB - structural collapse detail. . . 95
5.32 Honda Accord ODB - firewall deformation . . . 96
5.33 Honda Accord MPDB - acceleration and section forces . . . 98
5.34 Honda Accord MPDB - simulation snapshot . . . 99
5.35 Honda Accord MPDB - structural collapse detail . . . 100
5.36 Honda Accord MPDB - firewall deformation . . . 101
5.37 Honda Accord MPDB - A pillar and windscreen deformation . . . 102
5.38 Honda Accord - energy content comparison . . . 103
5.39 Honda Accord - ODB firewall deformation . . . 104
5.40 MPDB . . . 104
5.41 Honda Accord - MPDB firewall deformaton . . . 104
5.42 Honda Accord - front deformation comparison . . . 105
5.43 Chevrolet Silverado ODB - simulation snapshot . . . 108
5.44 Chevrolet Silverado ODB - structural collapse detail. . . 108
5.45 Chevrolet Silverado ODB - acceleration . . . 109
5.46 Chevrolet Silverado ODB - section forces . . . 109
5.47 Chevrolet Silverado ODB - cross section location . . . 110
5.48 Chevrolet Silverado ODB - overall cabin deformation . . . 111
5.49 Chevrolet Silverado ODB - driver side firewall intrusion . . . 111
5.50 Chevrolet Silverado ODB - passenger side intrusion . . . 112
5.51 Chevrolet Silverado MPDB - simulation snapshot . . . 114
5.52 Chevrolet Silverado MPDB - structural collapse detail. . . 114
5.53 Chevrolet Silverado MPDB - acceleration . . . 115
5.54 Chevrolet Silverado MPDB - section forces . . . 115
5.55 Chevrolet Silverado MPDB - driver side firewall intrusion . . . 116
5.56 Chevrolet Silverado MPDB - passenger side intrusion . . . 116
5.57 Chevrolet Silverado - ODB vs MPDB energy content . . . 118
5.58 Chevrolet Silverado - crash pulse comparison . . . 119
5.59 Chevrolet Silverado - velocity trend comparison . . . 119
5.60 Chevrolet Silverado - main rail deformation comparison . . . 120
5.61 Chevrolet Silverado - section forces comparison . . . 121
5.62 Chevrolet Silverado - firewall intrusion comparison. . . 122
5.63 U Model ODB - simulation snapshot . . . 125
5.64 U Model ODB - structural collapse detail. . . 125
5.65 U Model ODB - acceleration . . . 126
5.66 U Model ODB - section forces . . . 126
5.67 U Model ODB - cross section location. . . 127
5.68 U Model ODB - firewall intrusion . . . 128
5.69 U Model ODB - cabin deformation and interior intrusion . . . 129
5.72 U Model MPDB - acceleration . . . 132
5.73 U Model MPDB - section forces . . . 133
5.74 U Model MPDB - firewall intrusion . . . 134
5.75 U Model MPDB - cabin deformation and interior intrusion . . . 135
5.76 U Model - energy content comparison . . . 137
5.77 U Model ODB - structural damage . . . 138
5.78 U Model MPDB - structural damage . . . 139
5.79 U Model - crash pulse comparison . . . 139
5.80 U Model - velocity comparison . . . 140
5.81 U Model - RHS section force comparison . . . 141
5.82 U Model - LHS rail section forces comparison . . . 141
5.83 U Model - firewall intrusion comparison. . . 143
5.84 U Model - interior intrusion comparison. . . 144
6.1 Toyota Yaris - barrier deformation in area of interest . . . 150
6.2 Toyota Yaris - overall barrier deformation . . . 150
6.3 Toyota Yaris - barrier and vehicle acceleration comparison . . . 151
6.4 Toyota Yaris - barrier and vehicle velocity comparison. . . 151
6.5 Honda Accord - barrier deformation in area of interest . . . 153
6.6 Honda Accord - overall barrier deformation . . . 153
6.7 Honda Accord - barrier and vehicle acceleration comparison . . . 154
6.8 Honda Accord - barrier and vehicle velocity comparison . . . 154
6.9 Chevrolet Silverado - barrier deformation in area of interest . . . 156
6.10 Chevrolet Silverado - overall barrier deformation . . . 156
6.11 Chevrolet Silverado - barrier and vehicle acceleration comparison . . . 157
6.12 Chevrolet Silverado - barrier and vehicle velocity comparison . . . 157
6.13 U Model - barrier deformation in area of interest. . . 159
6.14 U Model - overall barrier deformation . . . 159
6.15 U Model - barrier and vehicle acceleration comparison. . . 160
6.16 U Model - barrier and vehicle velocity comparison . . . 160
6.17 Comparison between analysed vehicles . . . 161
A.1 Toyota Yaris FWRB - simulation energy . . . v
A.2 Honda Accord FWRB - simulation energy . . . vi
A.3 Chevrolet Silverado FWRB - simulation energy . . . vi
A.4 U Model ODB 56km/h - simulation energy . . . vii
B.1 Toyota Yaris FWRB - LHS comparison . . . ix
B.2 Toyota Yaris FWRB - RHS comparison . . . x
B.3 Toyota Yaris FWRB - top view comparison. . . xi
B.4 Toyota Yaris FWRB - three quarter view comparison . . . xii
B.5 Toyota Yaris FWRB - underbody comparison . . . xiii
B.6 Honda Accord FWRB - LHS comparison . . . xiv
B.7 Honda Accord FWRB - RHS comparison . . . xv
B.8 Honda Accord FWRB - top view comparison . . . xvi
B.9 Honda Accord FWRB - three quarter view comparison . . . xvii
B.10 Honda Accord FWRB - underbody comparison . . . xviii
B.11 Chevrolet Silverado FWRB - LHS comparison . . . xix
B.12 Chevrolet Silverado FWRB - RHS comparison . . . xx
B.13 Chevrolet Silverado FWRB - top view comparison . . . xxi
B.14 Chevrolet Silverado FWRB - three quarter view comparison . . . xxii
B.15 Chevrolet Silverado FWRB - underbody comparison . . . xxiii
B.16 U Model ODB 56km/h - RHS comparison . . . xxiv
B.17 U Model ODB 56km/h - underbody comparison . . . xxv
Chapter 1
Introduction
1.1 The issue of representative crash testing
In the 28 EU member states, road accidents are still a major cause of death, with over 25 000 people being killed every year and a much greater number suffering serious injuries [18]. More than 50% of these fatalities have been shown to be car occupants, involved in either single or multiple vehicle crashes [19]. This clearly poses an important problem firstly for social aspects, but also for economic reasons, as road deaths are a significant expense for european countries both in terms of tangible costs and human costs . For these reasons, throughout the last three decades, national and international institutions have been focusing on creating legislative and consumer tests in order to drive the manufacturers to improve the safety of their products, both to protect the occupants in the event of a crash, and to develop and install systems that could prevent the occurrence from happening in the first place. Among these, one of the projects that has had the most impact has been the European New Car Assessment Programme - Euro NCAP - which completes non legislative tests on the most popular vehicles in the european circulating fleet, assigning a star rating based on the performance shown in a set of standardised procedures. Since its introduction in 1996, the total death toll has been reduced by roughly a quarter, despite the large increase in traffic volume in the same timeframe. A substantial portion of the merit has to be attributed to this institution. In fact, after the first ten years of work, several studies that cross checked test results and accident data proved a positive correlation between occupant protection capabilities and high star rating [20], while vehicles with only 2-3 stars showed worse real life performance. Nonetheless, very relevant differences are present between the results seen in consumer testing procedures and real car-to-car impacts. In fact, the main aim of Euro NCAP was not the prediction of real life behaviour, but the assessment of best practices for specific car models and the overall circulating fleet, mainly due to the boundaries imposed by laboratory testing which could not give a comprehensive overview of the complex phenomena encountered on the roads. To fulfill this goal, the frontal testing procedure that was chosen since the very beginning was very similar to that used for legislative testing: a partial overlap frontal impact against a fixed deformable barrier, or Offset Deformable Barrier test (ODB).
Without diminishing the positive impact that the implementation of such procedure has had on the circulating fleet, which has become much safer, it was pointed out that the
use of the ODB test both in legislative and consumer crash testing has never addressed some important issues, while also being cause to others. Mainly, the research conducted by european entities and projects, such as ADAC and FIMCAR, has shown that the discrep- ancy between the rating obtained in laboratory testing and the performance in car-to-car accidents can be attributed to issues of compatibility between vehicles. This involves differ- ences in mass, front end structural design and structural rigidity between the two parties, all aspects that have never been considered in safety ratings due to the difficulty of as- sessment. In addition, the nature of the ODB procedure has worsened the situation by forcing vehicles with higher mass to increase their front end rigidity, due to the much larger amount of energy involved in the impact compared to light vehicles. As a result, in 2009, the accident analysis conducted by ADAC [19] has demonstrated that the probability of serious or fatal injuries in a two car collision were still double if the occupant was in a light vehicle, rather than in one with large mass.
For these reasons and thanks to technological advancements, Euro NCAP has recently decided to revolutionise its frontal impact testing protocol, in order to implement a proce- dure which represents the real life scenario more closely and also gives the ability to rate the compatibility level of the vehicles, adding to the overall score a number of considerations related to partner protection. The new test, which will be intruduced in 2020, involves the impact with a moving trolley of a mass representing the average vehicle ciruclating on EU roads. The trolley will have a different deformable element mounted on its front, which will replicate more effectively the structural stiffness of an idealised opponent car and also allow the evaluation of partner protection parameters based on its own deformation. Hence, it was denominated the Mobile offset Progressive Deformable Barrier test, or MPDB.
This change represents an important step forward for vehicle safety, as it will force OEMs and design companies to modify to a great extent current structural design trends in order to obtain the same level of performance that was seen in the past in terms of Euro NCAP rating. For this reason, it is important to start the process of understanding what are the effects of the new test on different vehicle classes and chassis design philosophies, as it represents an important first step towards the definition of the direction that will have to be followed in future vehicle design projects.
1.2 Aims and objectives
The overall aims of the study are to gain a deeper understanding, through the use of math- ematical modeling and simulation, of the effects caused by the new Euro NCAP procedure on passive safety performance for a sample of vehicles, selected in order to evaluate the dependency of these effects on the mass of the vehicle and on chassis architecture. Fur- thermore, a secondary aim is to also understand how these vehicles, designed either today or in the recent past, would perform if partner protection was to be considered, as it will be after 2020.
The objectives set out to achieve these aims are:
• to select a number of suitable mathematical models representing a small number of vehicles with different masses and chassis construction;
1.3 – Report structure
• to verify their representativeness by conducting a correlation study between the results obtained in simulation and those coming from physical tests;
• to perform a comparative study, simulating both the current Euro NCAP Offset Deformable Barrier procedure and the new MPDB test, to understand differences and criticalities between the two;
• and finally to assess their performance in terms of partner protection using the results from the MPDB simulations and implementing an analysis methodology that could effectively be used in the consumer testing rating protocol.
1.3 Report structure
This report is subdivided into seven chapters, five of these contain the research and findings of the work, while the initial and final ones are dedicated respectively to the introduction and overall conclusions. The main body of the report is structured as follows:
• the second chapter is aimed at providing a historical review of the evolution of the Euro NCAP programme, from its beginning to the future plans for 2020 and 2025, by analysing the existing literature. In order to have a clear view of the details of the two tests that will be compared in the results section, the procedures and parameters of the Offset Deformable Barrier and Mobile offset Progressive Deformable Barrier tests are presented. In conclusion, a comment on the importance of mathematical modeling and simulation for the vehicle design process and for passive safety is given;
• in the third chapter, the tools and methods utilised to complete the objectives of the study are reported. The initial decisions to be made regarding the choice of softwares to be selected for pre processing, solver and post processing are explained, together with a brief comment on their use. Next, a description and analysis of the four vehicle models adopted for the completion of the comparative study is shown, while the official models of the ODB and MPDB barriers created by the software provider are presented. In addition, the schedule of conducted tests is included for reference;
• the fourth chapter details the work performed to understand the level of correlation between the CAE models used and their physical counterparts. Initially, reasoning is given about the importance and value of this portion of the work, while in the second section the methodology to perform such a study is explained. Finally, the results of the analysis for each vehicle model are reported, together with conclusions regarding the effect of the findings on the following parts of the project;
• in the fifth chapter, the results obtained from the simulations of ODB and MPDB tests are presented. First, the models’ structural performance is analysed in terms of the ODB test, to assess their strengths and weaknesses on the current Euro NCAP testing protocol. Next, the MPDB test is analysed and finally a comparison between the two tests is completed. In the final section, conclusions on the findings are drawn;
• lastly, in the sixth chapter the work completed in order to gain an understanding of the partner protection level of the tested vehicles is reported. A brief description
of the procedure that the study is based on and of the methodology used to apply it in the simulation environment is given. The results of barrier deformation and acceleration obtained from the MPDB simulations of the four vehicles utilised are then analysed in order to highlight criticalities and positive achievements. Finally, the performance of the four vehicles is compared to draw conclusions on the findings.
Chapter 2
State of the art
In this chapter, a review of the evolution of the Euro NCAP consumer testing programme is given through the analysis of the existing litereature. In addition, the procedure of the current Offset Deformable Barrier test is presented, with details regarding both the setup of the vehicle, the parameters of the test and those of the barrier. The same is presented for the new Mobile offset Progressive Deformable Barrier test in order to have a clear picture of the changes before diving into the details of the results. Finally, a brief comment on the importance of crash test simulations in industry, especially in relation to passive safety, is given.
2.1 The evolution of Euro NCAP
The Euro NCAP (i.e. European New Car Assessment Programme) is a consumer crash testing programme established in order to provide to end users a realistic assessment of the safety level of new vehicles on the european market [21]. The mission of the organ- isation is to spread knowledge about effective vehicle safety, hence utilising the power of public opinion to push automotive manufacturers to improve safety systems beyond the homologation requirements. In turn, this directly affects the amount of lives saved in real world crash situations [3]. The method to achieve this objective has been based since the very beginning on a star rating system, calculated from the results of a range of tests, but simple enough in its end result to be comprehended by the entire consumer base.
The work of Euro NCAP originated in 1996, from the joint efforts of the Swedish National Road Administration (SNRA), the Federation Internationale de l’Automobile (FIA), the International Testing and the UK’s Department for Transport [3]. The initial idea was to move on the line of work of the National Highway Traffic Safety Association (NHTSA, USA) New Car Assessment Programme developed in 1979, where vehicles were tested in a full frontal impact with a rigid wall at 56km/h, although implementing the different testing strategies that had been developed in Europe. Namely, since the beginning the tests included the frontal offset test with ODB barrier at 64km/h, the side impact test with MDB barrier at 50km/h and the pedestrian impact tests for leg, upper leg and head.
All tests used in the original round of the programme were based on the developments by the European Enhanced Vehicle-safety Committee (EEVC) for legislation, with the
exception of a higher impact speed for the ODB, raised from 56 to 64km/h [5].
During the first two years of activity, four rounds were completed, starting from the 7 superminis of round one, moving on with a larger batch of family cars for round two, small family cars in round three and executive vehicles in round four. The selection of the cars was based on the highest selling version of the chosen models in the european market; although it is known that different body types, engine sizes and transmission types can influence crashworthiness, Euro NCAP did not aim at testing all vehicles on the road, which would be unrealistic and extremely resource consuming, but at providing information regarding the most significant variants for the road-going fleet. Furthermore, vehicles were only tested in the most basic configuration in terms of safety, meaning that optional equipment offered by the manufacturers at a higher price was not taken into account for the tests funded directly by Euro NCAP. The possibility was given to the manufacturers to fund additional tests in case a certain vehicle was not selected to be in the batch, or to prove the effects of additional safety equipment offered, or to have re-runs whenever an updated version of the vehicle was to come out. In order to ensure unbiased results, the tested vehicles were purchased from normal dealers anonymously, so to minimise the possibility of being given non standard production cars [5].
The initial reaction from the automotive manufacturers was highly critical, as it was believed that it would be impossible to achieve high ratings in the three tested categories due to the very strict evaluation system in place. Soon, however, the tide changed and OEMs started understanding the possible marketing advantage that would have come from scoring a high rating in the test, as public opinion was paying more and more attention to the published results. Not long after the first tests, vehicles were improved greatly and started scoring full marks in all fields, giving the first hint that the method enacted by Euro NCAP was in fact working. At this point, several manufacturers started offering most of the available safety equipment as standard and set the aim of reaching a four star rating as a primary design goal for new models [5].
The first major change to the testing procedures was implemented in 1999, when the pole test was introduced. The procedure was taken directly from the US side impact standard, with the only modification being the introduction of the EUROSID-1 dummy.
The decision behind this addition was based on the fact that, due to the Euro NCAP side impact test, manufacturers started introducing measures to reduce head and thorax injuries. However, the standard side impact test did not guarantee a consistent contact with the head of the dummy, hence no significant measure of the effectiveness of the new systems was present. Furthermore, data about road accidents showed that despite the relatively low number of accidents including an impact with a pole, the percentage of serious injuries or deaths caused by this kind of accident was incredibly high. With the new test in place, the scoring system was updated and the number of stars was increased from four to five, and the first vehicles to achieve a five star rating were tested in 2001.
As the years of activity increased, the work of Euro NCAP kept growing consistently, with two batches of tests per year, each containing an ever increasing number of vehicles, as a result of more national stakeholders taking part in the funding and a higher and higher commitment by OEMs. Together with this trend of increased testing volume, the range of areas taken into consideration in the star rating also widened. Amongst all the modifications introduced, the most significant were:
2.1 – The evolution of Euro NCAP
• in 2002, a score for seatbelt reminders was introduced in order to force manufacturers to, at the very least, include intelligent reminders for seatbelt usage for the driver;
• in the same year, the organisation noticed a lack of advancement in the field of pedestrian protection, which was at the time little regarded by the manufacturers, hence a modification to the scoring system for pedestrian protection was put in place, again to push OEMs to take the issue more seriously;
• in 2003, the first child protection rating was introduced, to make sure that vehicle producers took responsibility for the implementation of child restraint systems with the overall structure of the car, including ISOFIX structures as standard and giving customers a range of approved child seats by liaising with the aftermarket industry [3].
At the end of the first decade of testing, the results were very promising, as the amount of vehicles that started achieving a five star rating for adult occupant protection was very high, as highlighted in Figure 2.1. The amount of vehicles with a low star rating kept decreasing, while it was clear that the industry was reacting to all the changes introduced by Euro NCAP to obtain the wanted number of stars.
Furthermore, the results shown in Figure 2.2 clearly show how the modification to pedestrian rating enacted in 2002 drove the wanted outcome, with automakers taking the issue more seriously.
After the success of the first 10 years of work, 2009 was a meaningful milestone for Euro NCAP, as a great deal of modifications was introduced to the rating system. The main drivers for this were the rise of electronic driver assists and crash prevention technologies, not accounted for in the rating system at the time. In addition, the fact that the high number of vehicles achieving a five star rating was leveling the ground and the interest in the results was diminishing. The overhaul of the rating system consisted in the change from three different star ratings to one overall rating, calculated in a more complex way, in order to guarantee that manufacturers could not just achieve positive results in one field and mediocre performance in other less marketable ones. The new rating promoted heavily the use of fully integrated safety systems, comprising of high level technologies in
Figure 2.1: Progress of adult occupant star rating 1997-2007 [3]
Figure 2.2: Progress of pedestrian protection star rating 1997-2007 [3]
both passive and active safety for adult occupants, children and pedestrians. In this way, it would be easier to discriminate between vehicles performing optimally and vehicles still needing improvements in certain areas, as this would be reflected in a lower overall star rating. In terms of technology, great focus was put on driver assists, the main ones being:
Electronic Stability Control (ESC) Introduced in 2005 as a recommendation by Euro NCAP, in 2009 its integration became part of the rating and in 2011 a functional test was added;
Speed assistance systems (SAS) In 2009, the implementation of manually set speed limitation systems was included in the rating, with incentives for more advanced systems that could be set on the go. Later in 2013, with the advent of intelligent systems for speed limit detection and active assistance, the protocol was updated to include such technology as well;
Autonomous Emergency Braking (AEB) Forward collision warning and AEB sys- tems have represented the biggest revolution in active safety since the introduction of ESC, for this reason, both high speed and low speed AEB systems have been included in the star rating since 2014.
With regards to the field of passive safety, central interest of the work portrayed in this report, it was noted that that the most revolutionary improvements occurred already during the first decade of Euro NCAP’s work. The front structures have been gradually improved with the design aim of minimising intrusions in the passenger compartment during the 64km/h ODB test and the level achieved in 2015 was extremely satisfactory for the great majority, if not all, of the tested vehicles. Furthermore, great structural improvements also impacted the side crash occupant protection, which at this moment has reached a high level in terms of Euro NCAP rating. However, Euro NCAP is willing to push the boundaries of development even further, not only by taking into account the latest active safety technologies, but also asking for an extra effort in the area of passive safety to make sure that the structures are designed for impacts that are as similar as possible to real life situations. For this reason, the new Euro NCAP Roadmaps have been devised, giving to the industry an insight into what will be asked in the future by consumer crash testing procedures.
2.1 – The evolution of Euro NCAP
2.1.1 Euro NCAP future Roadmaps
In order to respond to the changes in technology, road accident data and to push the industry to go beyond the status quo, Euro NCAP has, at more or less regular intervals, published their vision and outlook for the future requirements they would be setting in the so called "Roadmaps". In 2015, before the ending of the period concerned by the previous document, set for 2016, the committee published the "2020 Roadmap", where the main objectives and changes for the following years were highlighted. As stated in Chapter 1, the main objective for the near future is to reduce even further the number of casualties and serious injuries on the road, and the way to reach this goal is still quite long. Several changes regarding technological innovation in terms of active safety have been set up for 2020, starting from the update of procedures for AEB, SAS and seatbelt reminders, to the inclusion of procedures for lane departure assists, speeding and impaired driving avoidance, and semi autonomous driving. However, the most substantial change to test procedures for passive safety since the introduction of the pole test has also been included. After 20 years of activity, the Offset Deformable Barrier test procedure at 64km/h will be finally retired. The test procedure will be substituted by the Mobile Offset Deformable Barrier test at 50km/h, with the aim of improving road accident representativeness and begin the very important consideration of compatibility and partner protection [22]. Furthermore, the second challenge will concern the side impact procedure, as a modification to the actual rating will enable the assessment of far side occupant protection for driver and passengers. Finally, the current Hybrid-III anthropomorphic test device will be replaced by the more recently developed THOR Advanced mid-sized male device, which enables a higher level of biofidelity and an unparalleled performance in terms of instrumentation and data acquisition technology [23].
2020 ROADMAP
EUROPEAN NEW CAR ASSESSMENT PROGRAMME March 2015
EURO NCAP 20/25 ROADMAP
1
Euro NCAP 2025 Roadmap
I N P U R S U I T O F V I S I O N Z E R O
Figure 2.3: Euro NCAP future Roadmaps
Due to the pace at which automotive manufacturers are progressing in the fields of au- tomated driving, active safety assistance and crash testing virtual simulation, Euro NCAP has already set the goals and directions for the more distant future by publishing in 2018 the "2025 Roadmap". The objective set for the future is to achieve the "vision zero", the complete elimination of road casualties. In the pursuit of this goal, the focus is set on primary and secondary safety, with increased focus on driver monitoring, autonomous aids and V2X communication systems for the former, and rear-end collision protection, pedes- trian and cyclist safety for the latter. In addition, for the first time in Euro NCAP’s history, tertiary safety is introduced: Child Presence Detection systems to solve the prob- lem of children forgotten inside vehicles will be taken into consideration, while an initial assessment at technologies to aid extraction from crashed vehicles will be carried out [24].
Overall, Euro NCAP has been constantly updating and looking to include new aspects in its rating, in order to maintain it pertinent, meaningful and most of all useful for road users. The future shaped by the decisions that have been made in the last few years continues on that path, with the ambitious objective of continuing to push the industry until no more fatalities occur on european roads.
2.2 From ODB to MPDB
Having briefly discussed the advancements completed in the first 20 years of Euro NCAP’s path and the ideas put in place for the short and mid term future, focus must be now placed on the matter of most interest to this report: the substitution of the outdated Offset Deformable Barrier test procedure with the newly developed Mobile Offset Deformable Barrier procedure. In order to understand to the full extent the importance of this change, the details of both tests together with their analysis is reported in the following sections.
2.2.1 The Offset Deformable Barrier test
The Offset Deformable Barrier test has been designed by the European Experimental Vehicles Committee (EEVC) in 1994 [4], with the aim of implementing an additional test to the full width rigid wall test used as european legislation at the time. Several studies regarding real world impacts highlighted the issue of high levels of injuries and mortality in frontal car-to-car crashes, which was deemed to be caused by contact between the occupant’s body and the vehicle structures due to high levels of intrusion . The discrepancy between the results in full width rigid wall tests and actual road accidents was found to be due to the different kind of loading suffered by the frontal structures of the vehicle: the decreased amount of overlap in the real scenarios had the effect of loading only one of the two sides of the vehicle, hence forcing only half of the structure to absorb all impact energy, thus forcing the cabin to deform to dissipate the residual energy; in addition, the perpendicular surface and extremely high deceleration rate experienced by the face of the front structures impacting against a rigid wall ensured that the collapse of said elements followed the desired buckling sequence, with stiffer structures absorbing most of the energy. This is very different from the real scenario, where the vehicle is impacting against a deformable object (partner vehicle), which does not ensure either a flat, stable surface to load, nor such high decelerations to the front structures. The result
2.2 – From ODB to MPDB
of this occurrence is the inability to collapse the stiffer structures that hence get pushed back in the vehicle as the weaker components fail [25]. Furthermore, in full width rigid barrier tests the engine tends to undergo very high accelerations, which equate to large energy absorption when the block impacts with a stiff firewall. In road impacts, such rigid structures do not exist and the engine cannot be loaded in the same way, and accident analysis had showed that generally a more realistic scenario involves the engine moving sideways and loading the firewall only partially.
For these reasons, the EECV working group designed a test with a limited overlap, set at 40% with the goals of loading mainly one side of the front crash structure and avoiding high engine loading, hence creating a more realistic condition. With regards to the impact face, a deformable barrier was required, aimed at reproducing the softer and more complex shape of a partner vehicle. Hence, several tests were carried out and it was concluded that the most suitable design would be a 450mm deep, 50psi aluminium honeycomb block with a smaller 250psi aluminium block attached to the lower part of the front face, as show in Figure 2.5. This barrier was denominated "normal element with bumper" [4]. In terms of positioning, the deformable face was placed on an "infinitely" rigid block at a height of 200mm from ground level.
Figure 2.4: ODB Normal element with bumper [4]
Regarding the impact speed, at the time of design of the procedure the decision was made to shape the test in order to reproduce with high fidelity a 50km/h car-to-car crash between identical vehicles with partial overlap. For this reason, the final test speed was set
to 56km/h, although it was already clear at the time that this might be an underestimation, as a speed above 60km/h would have been more representative of the harshness of the majority of real world crashes. Finally, this test represented the first instance in which anthropomorphic test devices were used: the Hybrid III 50th percentile male dummy was chosen in order to record biomechanical parameters and set limits to loads, accelerations and deformations of body components.
Figure 2.5: Car-to-car frontal impact speed and serious or fatal casualties [5]
Resulting from the work of EECV WG 11, the UNECE-94 regulation was set up for vehicle homologation in europe [26]. When Euro NCAP decided what test to include, the choice was made to take the specifications of UNECE 94 in all their declinations. The only change that was made regarded the test speed, set, as mentioned above, at 64km/h.
Based on the accident data analysis reported by EECV, it was found that an impact speed of 55km/h in a real life crash would replicate around half of the serious or fatal injuries;
at this point, comparative tests were conducted on a medium sized family car and it was found that an ODB test at 64km/h would be around the same severity of a car-to-car impact at 55km/h, due to the amount of energy absorbed by the deformable barrier [5].
Test specifications
The full specifications of the test parameters and barrier construction are here reported. In terms of added mass to the kerb weight of the vehicle, all fluids are topped up to standard running condition, the fuel tank is filled with water (or equivalent) to 90% of its capacity in terms of mass of fuel, and 36kg of ballast are added to the luggage compartment. In addition, the two Hybrid III ATDs placed in the front seats have a mass of 88kg each, while in the rear seats the Q6 and Q10 child dummies have a mass of 23kg and 36kg respectively.
The child dummies must also be placed on the child restraints recommended by the OEM;
if these are not available, an additional mass of 7kg for the Q6 and 2kg for the Q10 must be included [27].
The barrier is constructed with several layers: a main aluminium honeycomb block with a crush strength of 0.342 M pa, an aluminium bumper element with crush strength of 1.711 M pa, an aluminium backing sheet, cladding sheet and bumper facing sheet. The different elements are combined using a specific adhesive bonding procedure [26]. The barrier must be then fixed to a rigid block with a minimum mass of 7 ∗ 104kg and the attachment geometry should be such that, during the impact, the vehicle never comes into contact with the rigid block.
2.2 – From ODB to MPDB
Table 2.1: ODB test specification Vehicle velocity 64km/h
Overlap level 40%
Barrier type ODB
Front ATDs Hybrid III 50th Male [88kg]
Rear ATDs Q6 [23kg]- Q10 [36kg]
Child restraint mass 7kg (Q6) - 2kg (Q10)
Fuel equivalent mass 90% fuel tank capacity by mass
Luggage mass 36kg
Figure 2.6: ODB test infographic [6]
Figure 2.7: Physical model of ODB [7]
Criticism and issues of the ODB test
Since its introduction, more than 20 years back, the use of ODB test has given the initially desired results, driving OEMs to improve vehicle structures to higher and higher levels.
However, many aspects of the test have been criticised and many shortcomings have been identified, both when it was originally design and through the years. At the time of its implementation, it was already clear that the test would not allow the evaluation of the effect on the partner vehicle, as this was not in the objectives of the working group. The
Table 2.2: ODB construction specifications
Height Width Depth Material Crush strength
[mm] [mm] [mm] [M pa]
Main block 650 1000 450 Al 3003 0.342
Bumper element 330 1000 90 Al 3003 1.711
Backing sheet 800 1000 2 (thickness) Al 5251 -
Cladding sheet 1700 1000 0.81 (thickness) Al 5251 - Bumper sheet 330 1000 0.81 (thickness) Al 5251 -
barrier dimensions and material was not designed for this kind of assessment, while the test procedure itself could not assess phenomena of the likes of under- and over-riding, issues that have become more relelvant as the level of occupant protection has reached satisfactory levels [28].
Furthermore, as stated above, the ODB test was designed to replicate a car-to-car impact with same vehicle; this poses a big issue in terms of representativeness of real life scenarios, as a light vehicle as a high probability of impacting with a heavier vehicle and vice versa. In the case of a crash between a light and a heavier vehicle, the one with lower mass will experience higher loading due to the law of conservation of momentum [29], hence the structures will have to absorb a level of energy outside of the design range imposed by the ODB test. This results in levels of cabin intrusion substantially higher than tested and also in higher accelerations, again due to the conservation of momentum.
The same issue is reflected in an opposite manner for heavy vehicles in the test. Due to the fixed, and relatively low, amount of energy that the barrier can absorb, the energy that the structure of a heavy vehicle has to absorb is proportionally larger than that of light cars. This lead to a trend of increased stiffness of front structures in already aggressive, large mass vehicles, causing even more compatibility issues when a real impact occurs, while also affecting overall vehicle mass with all its unwanted consequences on performance parameters. Studies have also found that more in general, for all types of vehicles the ODB test has driven a high increase in stiffness of the front crash structures, hence creating a problem for compatibility with partners and adaptability to different impact scenarios [30].
Finally, the soft and shallow ODB barrier has led to a negative design trend in industry:
most vehicles nowadays have crash structures which puncture the barrier on purpose, in order to exploit the rigid back plate. This, in a way, makes the test more similar to a partial overlap rigid wall test, hence the type of loading undergone by the crash structures becomes more perpendicular in direction and more abrupt in acceleration, ensuring appropriate collapse in a simpler way. This is clearly an overexaggeration, as the barrier, while being punctured, still does perform partially its function of simulating a partner vehicle. The point that has been raised is, however, that the technological level reached has allowed the implementation of a better approximation of real life conditions compared to what was possible when ODB was designed [31].
2.2 – From ODB to MPDB
2.2.2 The Mobile Offset Deformable Barrier test
Due to the issues stated in the previous section, a new test was studied and devised through the last 15 years with the aim of substituting the UNECE-94 test for homologation to improve the level of safety directly from a legal standpoint. However, an agreement has not been found yet for application on such a large scale, even after the extensive finalisation work by the ADAC MPDB [19] and FIMCAR X projects [32]. The core area to tackle for this new test was the evaluation of compatibility in the broad sense of the term, hence taking into account the effects of the mass of the vehicle, the front end rigidity, the front end structures design, both in regards to occupant protection and also in terms of partner protection, creating a scenario of higher fidelity to road accidents [19].
The development of the new test took shape from the joint agreement of several organ- isations, both from Europe and USA, that the most suitable procedure for the evaluation of compatibility would be a mobile barrier test. The underlying idea that drove the design of the ODB test was not modified, as the accident analysis showed that the relevance of an offset test to simulate a car to car impact at a speed between 50 and 60 km/h is still valid, and through the years several studies on possibilities for mobile frontal offset tests were carried out. From this starting point, the FIMCAR X project, reuniting the major european organisations and test labs, conducted 15 full scale tests on vehicles of different sizes to determine the parameters of the protocol to be followed, mainly in terms of test speed, barrier mass and overlap. The procedure for of the mobile barrier test involved a test vehicle and a mobile trolley with a front mounted barrier face; the two bodies are positioned facing each other and are launched at the same, and opposite, velocity towards one another. The alignment of the barrier face with the front of the test vehicle will be equal to the desired offset. The study included superminis, small and medium family cars and also SUVs of different sizes, with a minimum vehicle mass of 1000 kg and a maximum of 2200kg and arrived at the conclusion of a proposal for a new test procedure [33], later to be adopted by Euro NCAP [34], which decided to update its roadmap to put priority on its implementation.
The procedure described above, as opposed to the ODB test, does not intend to replicate for every vehicle a crash with a similarly sized opponent, but an impact with a car of set mass driving on the road. In the case of Euro NCAP and the referenced european project, this mass corresponds to the average of the circulating fleet on european roads.
However, the same test could be used, by changing the mass parameter, to replicate other populations of vehicles in different parts of the globe, hence making this procedure an appropriate candidate for worldwide standardisation and harmonisation of tests, while also being extremely robust to future changes without the need for complete redesign and extensive validation tests [35].
As it was far out of the scope of the project to design a new barrier element and a new trolley, the Progressive Deformable Barrier (PDB) face developed in the VC-Compat project [36] and used in FIMCAR V [37] project for the WP2 offset test was selected, while the trolley utilised in side impact tests was deemed to be a suitable base. The design of this barrier element is based on having three different layers, a soft outer bloc, a middle one being capable of absorbing a much larger amount of energy and an inner layer with elevated crush strength. All in all, the PDB barrier is capable of absorbing a significantly higher amount of energy compared to the ODB element. In this way, the vehicle impacting
the barrier finds it considerably more difficult to puncture the whole depth and reach the rigid element of support, with the advantage of both reproducing in a closer way the front end structure of an idealised vehicle and of giving the possibility of assessing the damage inflicted to the barrier by the vehicle in a significative manner. The clear conclusion is that the assessment of the deformation can give a level of insight into the partner protection level of the tested vehicle.
The tests reported in the projects cited above included an assessment of speeds of 45, 50 and 56 km/h for both the vehicle and the impactor, with an overlap set at 50% and an impactor mass of either 1400kg or 1500kg. In both [19] and [32] it was found that a test speed of 56km/h resulted in deformation and acceleration pulses which were significantly higher than the reference car-to-car impact. The opposite occurred when a speed of 45km/h was utilised, hence the baseline test at 50km/h was adopted as the most suitable, as in fact it represents closely the reference test. The initial choice of barrier mass of 1500kg was driven by the results of the GIDAS accident study [29], which highlighted how front seat occupants of vehicles with mass lower than 1500kg are more likely of being seriously or fatally injured compared to occupants in vehicles heavier than the set threshold. However, for the final proposal of the procedure, the barrier mass was lowered to 1400kg, as it better represents the average mass of the compact vehicle category in Europe, which is the most widely sold [34]. Regarding the overlap level and direction of impact, the decision was again based on the German (GIDAS), French (LAB) and Swedish (VCTAD) accident data, which highlighted how the 12 o’clock direction would be absolutely relevant and a maximum overlap level of 75% would be representative of most occurrences of serious or fatal injuries in frontal impacts [34]. Finally, the 50% overlap level cited above was chosen.
During the development and after the release of the first drafts of the new procedure, a moderate number of research studies and tests have been completed. These have high- lighted how the MPDB is in fact capable of showing the shortcomings of the current standards for passive safety for lighter vehicles, derived from the utilisation UNECE-94 and Euro NCAP ODB, as the crash pulse severity was substantially increased as the whole crash event lasted between 30 and 50ms less than ODB [38], while in some cases their structures would not be able to guarantee the required levels of intrusion. With regards to vehicles with mass substantially higher than the trolley, the test severity is expected to be lower, which is again more representative of real car-to-car accidents and could lead to a modification of frontal structures’ stiffness that could diminish the aggressiveness towards lighter vehicles [35]. Other studies such as [39] have also demonstrated how on certain models with mass comparable or lower to the trolley the levels of intrusion could be far above the required limits, due to the modified behaviour of the crash structures. Finally, for vehicles with mass in the same range or slightly higher than 1400kg, the crash sever- ity was not substantially diminished and injury assessment values were still more severe than those obtained through ODB: even though the change in velocity is expected to be lower for vehicles with mass above 1400kg, the PDB structure is much stiffer and loads the structures in a considerably different manner [29]. The general image depicted by the current publications available is that of a high level of variability not only depending on mass, but also on the design of the front end of the vehicle which leads to either high crash pulses, high levels of intrusion or both, when the structural behaviour is incorrect.
2.2 – From ODB to MPDB
With regards to the second, but critical aspect represented by partner protection and compatibility assessment, studies have now been completed mainly with the aim of achiev- ing a proposal for the assessment method to be implemented, as reported by [33] and [19]. The procedure that will be adopted by Euro NCAP in 2020 is still unknown, as its publication is expected through 2019. The aim of the procedure will be to give an ob- jective evaluation of the aggressiveness of a vehicle with regards to its partner, favouring designs that consider the deformation and pulse inflicted to the opponent at a similar level of importance to occupant protection. In order to complete the evaluation, the general agreement is that a 3D scan of the deformed PDB barrier will be used and a quantification of the discrepancies in deformation between different areas of the honeycomb blocks will be the basis for the assessment: a vehicle that is capable of deforming the barrier in a homogeneous manner is far more likely to engage the front structures of the partner vehi- cle in a positive fashion, driving a deformation similar to a rigid block. Furthermore, the deformation impressed in the barrier can be analysed from a qualitative point of view, to investigate the aggressiveness of the frontal structures and their behaviour, while under or over-riding tendencies will also be highlighted. Only a few published projects have reported the results of the implementation of such procedures, such as [40] and [31], and all have shown extremely poor performance of the tested vehicles due to critical failures of struc- tural elements and interaction with engine block and wheels. This has already proved how the implementation of MPDB as a Euro NCAP standard will drive a heavy improvement in the design of safety structures, shifting the focus from strict self protection to self and partner interaction.
Test specifications
The test specifications published in the aforementioned reports by FIMCAR [32], ADAC [19] and Euro NCAP working group [34] are the most up-to-date drafts available at this time; however, it is recognised that the latest version reported by Volker [34] could very well be identical to the final specification which should be published in the near future, at least for the most part. For this reason, this draft is taken as the full specification for the work portrayed in this report and its details are highlighted in this section. Furthermore, the specifications for the barrier face and trolley have also been published as a draft, directly by Euro NCAP in 2017 [10]. Similarly to what has been stated for the test procedure, this draft will be taken as definitive for the purpose of this project.
Figure 2.8: MPDB test infographic [9]
The specifications of the test regarding the vehicle present few differences in terms of added mass: the THOR ATDs are placed in the front seat and their mass is slightly lower compared to Hybrid III dummies, setting the scale at 80 kg. The rest of the procedure in these terms remains unchanged, with Q6 and Q10 child dummies in the back seats, 90%
of fuel by mass in the fuel tank and 36kg of mass in the luggage compartment. As stated previously, the test speed has been set to 50 km/h both for the vehicle and for the trolley while the overlap has been fixed to 50%.
Table 2.3: MPDB test specification Vehicle velocity 50km/h
Overlap level 50%
Barrier type PDB
Barrier velocity 50km/h
Front ATDs THOR ADV 50thMale [80kg]
Rear ATDs Q6 [23kg]- Q10 [36kg]
Child restraint mass 7kg (Q6) - 2kg (Q10)
Fuel equivalent mass 90% fuel tank capacity by mass
Luggage mass 36kg
Regarding the barrier face, the Progressive Deformable Barrier is composed of three dif- ferent aluminium honeycomb deformable cores fixed one in front of the other, denominated A, B and C starting from the one closer to the trolley face. The blocks have the dimen- sions shown in Figure 2.9, with blocks A and C showing a homogeneous crush strength characteristic. The first honeycomb core (A) must have a strength between 1.540M P a and 1.711M P a when statically loaded in accordance with procedure NHTSA TP-214D, while for the third (C) the value must be between 0.308M P a and 0.342M P a when tested in the same manner. Block B is effectively the "Progressive" element in the barrier’s con- struction: its crush strength must be variable in accordance with compression, with a characteristic within the boundaries shown in Figure 2.10, when tested with a compres- sion rate of 100mm/min from 0 to 355mm. Additionally, the barrier comprises of a back mounting plate, three intermediate plates, a contact plate and a cladding sheet, as shown in Figure 2.11. The dimensions and material characteristics of all elements are reported in Table2.4. All components of the barrier are held together by a two-part polyurethane adhesive agent.
With respect to the mobile element, or trolley, the Euro NCAP specification sheet sets a total mass of 1400kg and the constructive constraint of no deformation after the impact with the vehicle. Front and rear track are fixed at 1500mm, while wheelbase must be equal to 3000mm. The centre of gravity of the barrier should be located on the vertical plane connecting the centres of the two axles, 1000mm behind the front axle and at a height of 500mm.
2.2 – From ODB to MPDB
Table 2.4: PDB construction specifications
Height Width Depth Material Crush strength
[mm] [mm] [mm] [M pa]
Block A 568 1000 90 Al 3003 1.540-1.711
Block B 568 1000 450 Al 3003 Progressive
Block C 568 1000 250 Al 3003 0.308-0.342
Backing sheet 720 1000 3 (thickness) AlMg2/3 -
Cladding sheet 720 1000 0.8 (thickness) Al 5754 - Intermediate sheet 566 1000 0.5 (thickness) Al 5754 - Contact sheet 566 1000 1.5 (thickness) Al 1050A -
Figure 2.9: PDB physical model [7] and dimensions [10]
Figure 2.10: Displacement-strength characteristic corridor of Block B [10]
Figure 2.11: Progressive Deformable Barrier physical model [7]
2.3 Crash test simulation in the vehicle design process
For the past 15 to 20 years, the vehicle design process has been highly influenced by the development of Computer Aided Engineering and Finite Element modeling and simula- tion, as the advancements in such fields, combined with the ever increasing availability and affordability of powerful computing hardware, have brought much higher flexibility to the design cycle. With regards to crash testing, the advent of simulation has impacted heavily the possibilities of car makers and researchers since the earliest stages of its intro- duction [41], as the possibilities to gather data from full scale physical tests has always been extremely restricted due to budget constraints. In fact, one if the main issues in the advancement of vehicle safety is the extremely high cost of physical tests, especially during the design process, when prototypes have to be crashed: it would be absolutely unfeasible to conduct a full scale physical test every time a modification is performed to vehicle structures, restraint systems, ATD positioning, test procedure and so on. This is even more valid when research projects are concerned and a high number of variations to the subject under study are undertaken. The evolution from linear static to non linear dynamic finite element modeling in commercial codes such as LS-DYNA and PAM-Crash
2.3 – Crash test simulation in the vehicle design process
has allowed engineers to improve their understanding in terms of full vehicle structural behaviour and crash dynamics [42], thanks to the possibility of visualising every step of the impact event and repeating simulations with design improvements and variations at an extremely high pace compared to the past. The evolution of both vehicle mathematical models and simulation codes has also enabled designers to take advantage not only of the qualitative assessment of simulated crash events, but also to utilise the resulting quanti- tative data regarding deformations, accelerations, forces and more as a close guideline to direct the vehicle design process. In fact, the simulation of crash events constitutes, nowa- days, one of the pillars of every design cycle: since the initial stages of the chassis design, simulations are preformed in order to verify if the direction taken by engineers is consistent with the expectations and objectives set for passive safety. As the process continues, the mathematical model of the vehicle becomes more and more similar to the finalised product, with every component eventually being represented with a high level of fidelity, both in terms of shape and mechanical-physical qualities. As a result, the simulations that derive from such detailed models become closely representative of the results that will be obtained in physical testing.
The level of accuracy and confidence with whom the full scale crash test simulations represent reality is difficult to be judged from a theoretical point of view, and the scarcity of publications on the matter does not allow to quantify with absolute certainty the repre- sentativeness of the results. However, the extensive use of such tools in industry and the verifications that have been performed through the years as several design projects devel- oped has ensured that using simulations as a research and design aid is extremely valuable and closely representative. Although the variability of the results in physical tests and the complexity of the calculation of full scale crashes are both considerable, the qualitative and quantitative data obtained in simulation gives information that is consistently in the same region of interest of the physical scenario. In conclusion, this proves how advanced simu- lation tools can be effectively utilised in order to perform research analyse to evaluate new scenarios, such as that of interest of this report, with a considerable amount of confidence that the results be in line with reality.
